TNA 102. Characteristics of the Spark Analogue Telephone Network Customer Interface. TNA 102: October 2017 DRAFT FOR COMMENT.

TNA 102 Characteristics of the Spark Analogue Telephone Network Customer Interface DRAFT FOR COMMENT Access Standards Spark Limited Wellington NEW ZEALAND October 2017 1

CONTENTS REFERENCES 2 FOREWARD 2 SPARK DISCLAIMER 3 1 SCOPE 4 2 NETWORK INTERFACE CHARACTERISTICS 5 3 DEFINITIONS 7 4 TRANSMISSION CHARACTERISTICS 12 5 SIGNALLING 20 6 D.C. LINE CONDITIONS 22 7 RINGING CHARACTERISTICS 25 8 SUPERVISORY SIGNALS 28 9 ANALOGUE ON-HOOK DATA TRANSMISSION 30 10 ANALOGUE CALLING LINE INDENTIFICATION PRESENTATION 34 11 VISUAL MESSAGE WAITING INDICATION 37 12 CUSTOMER SERVICE DELIVERY POINT PHYSICAL INTERFACE 38 13 SUMMARY OF DIFFERENCES BETWEEN INTERFACES DELIVERED BY DIFFERENT TECHNOLOGIES 39 REFERENCES TNA 151 :Telecom Network Transmission Plan PTC 107 : PABX External port interface requirements PTC 200 : Requirements for Connection of Customer Equipment to Analogue Lines May 2006 PTC 220 : Requirements for Private Voice Networks connected to PSTN/ISDN PTC 226 : Telecom Requirements for 2-wire 2 pin Sockets for Residential Use TCF Premises Wiring Cable Installers Guidelines for Telecommunication Services TCF Document: SIP ATA Standard for LFC Wholesale Service (Loose Coupling) version: 1.31 Date 26 March 2015 ITU-T Recommendation Q.552: Transmission characteristics at 2-wire analogue interfaces of digital exchanges ITU-T Recommendation G.168: Digital Network Echo Cancellers 2

FOREWORD This document describes the nominal characteristics at the customer Service Delivery Point (SDP) for analogue connections to the Spark voice network. Since the original publication of TNA 102 there have been significant changes to the network with analogue telephony now being delivered through a range of technologies in addition to a copper pair connecting directly to a NEAX digital TDM exchange. This Specification will cover the characteristics of the different delivery mechanisms which, while being broadly the same as the original NEAX based PSTN have some significant differences. SPARK DISCLAIMER Spark makes no representation or warranty, express or implied, with respect to the sufficiency, accuracy, or utility of any information or opinion contained in this Specification. Spark expressly advises that any use of or reliance on such information is at the risk of the person concerned. Spark shall not be liable for any loss (including consequential loss), damage or injury incurred by any person or organization arising out of the sufficiency, accuracy, or utility of any such information or opinion. This document describes the conditions encountered on the majority of Spark connections. It does not cover the extreme conditions that may arise on a small proportion of the total connections in the network. As an example, service in some rural areas may be provided by a combination of cable, transmission systems and/or radio systems which may not support all Spark services or terminal equipment functions. It must be stressed that the Spark Voice network is designed for telephony. While it generally supports basic data and facsimile transmission, these functions require specific network implementations and the performance is likely to vary from place to place depending upon how service is delivered. It must also be stressed that some of these interfaces are provided by other infrastructure providers and the parameters of the interface are can vary from one provider to another, and while Spark will endeavour to keep this document up to date, changes may occur from time to time where some interface parameters change ahead of this Specification. 3

1 SCOPE This document describes the analogue customer interface to the Spark voice network at the Service delivery point at the customer s premises. At this interface, analogue customer equipment such as a telephone or facsimile machine meeting the requirements of PTC 200 is normally connected. Historically, the service delivery point was simply connected back to the local telephone exchange by a cable. From the 1960s, rural customers could be served by either an analogue frequency division multiplexer or for very remote customers, a radio link (known as a country set) was used. These were relatively simple systems in that they simply monitored the electrical signals from one end and regenerated them at the other, in addition providing a two-way audio path. Signalling was relatively simple; loop disconnect for dialling outgoing calls and ringing for incoming calls, and supervisory functions simply the remote party listening to audio tones on the audio (speech) path. These analogue systems are no longer in use, and have been replaced by new digitally based systems. The main systems in use today are as follows: Direct copper cable pair connections to a Time Division Multiplex (TDM) telephone exchange (PSTN). Direct copper pair connection to a TDM multiplexor in a Chorus Cabinet. Direct copper pair connection to an IP line card in a Chorus Cabinet. Direct connection to an Analogue Terminal Adapter (ATA) incorporated in an Optical Network Termination (ONT) located in the customer s premises. Connection to an ATA incorporated in a 4G wireless terminal located in the customer s premises. Direct copper pair connection to a multi-access radio terminal. In addition to these interfaces, PBXs and Private Networks with Single Line Telephone (SLT) interfaces complying with PTC 220 section 5 can be classed amongst these interfaces. This Specification documents the customer interfaces of the above which allows customer equipment designers/manufacturers to supply equipment which can work satisfactorily with all of the interfaces. It should be noted that as the network evolves, the access technology used, and therefore the interface at the service delivery point, are likely to change as well. At this stage, the majority of customer connections are still in the first two categories which is described by the original version of TNA 102. This specification will specify each parameter, with any variations documented with the specification for that parameter. In addition, a summary of the differences from TNA 102 will be presented in Section 13 of this Specification. 4

2 NETWORK INTERFACE CHARACTERICS The Characteristics at the Service Delivery Point (SDP) for some methods of access cannot be precisely defined for every installation as there are a range of access line parameters which vary from customer to customer. This document will define the range of conditions likely to be encountered on the majority of lines. Figure 1 shows the functional blocks which are present in an analogue customer interface to a telephone network. Tones/Audio Message* Cable between network equipment port and customer equipment (0 to 6 km) Customer Service Delivery Point Network equipment 2-wire interface Ringing DC Feed 4/2 wire hybrid Loop Detector R-Pad Attenuation from the network to the customer T-Pad Attenuation from the customer to the network A D Balance Impedance A D 0 dbr point Control and network interface To Core network Tone Detector** * Tones fed to the customer interface may be carried in the audio path from else where in the network or may be generated in the interface. These tones are generally used for call progress indicators (busy tone, ringing tone etc ) but also include FSK Caller ID and message waiting indicators ** Tone detection may be performed at the interface (typically for packet based systems) or the tones may be carried back to the core network via the speech path (typically for TDM based systems) Figure 2.1 Functional Block diagram of an analogue customer interface Points to note are: Where the interface is remote from the core network, the connection to the core may be via copper, Fibre or Wireless, and may use TDM or packet techniques. In the case where the interface is part of the telephone exchange, the interface (Line card) is connected directly to the TDM bus within the NEAX switch. Where there is a copper cable access line between the customer and the network interface equipment, the objective has been to engineer the network so that the traffic weighted mean loss between the customer and the interface is 2.5 db. Where the interface equipment is located in the 5

customer s premises, the T and R pads in the interface are increased by 2.5 db to compensate for having a virtually zero length access cable. This specification is broken into 8 sections, defining in the interface as follows: 1. Transmission Characteristics of the audio path 2. Signalling from the Customer equipment to the network 3. d.c. characteristics 4. Ringing characteristics 5. Supervisory Tones 6. Analogue On-hook data transmission 7. Customer Service Delivery Point physical interface 8. Summary of differences between interfaces delivered by different technologies 6

3 DEFINITIONS 3.1 In general, definitions set by the International Telecommunications Union and published in the ITU-T Recommendations apply throughout this Specification. Nevertheless, some ITU-T definitions are not particularly informative for those unfamiliar with telephone engineering and therefore the following definitions are provided. Where necessary, these are supplemented by explanatory notes which elaborate on the formal wording. 3.2 Additional definitions are provided in all PTC Specifications. Nevertheless, some definitions are repeated in this document for ease of use and for explanatory purposes:- Analogue Terminal Adapter (ATA): is a network to customer interface as a standalone piece of equipment which provides all the network interface functions. It is usually located in the customer s premises. It is connected to the network core by fibre, wireless or DSL over copper. Also known as an FXS (Foreign Exchange Subscriber) interface. Called party: is the person or device receiving a call. Calling party: is the person or device initiating a call. Convergence: is a term used in connection with echo cancellation and is the process of developing a model of the echo path which will be used to estimate the circuit echo. See also "echo control device". Crosstalk: is any unwanted signal introduced into a line or equipment through coupling between one or more other lines or items of equipment not electrically connected. dbm: is the absolute power level in decibels (db's) relative to 1 mw. dbm0: is the absolute power level in decibels referred to a point of zero relative level (0 dbr). dbr: is the nominal relative power level in decibels referred to a point of zero relative level (0 dbr). Decadic: is the form of call initiation signalling which makes use of one or more timed disconnections of the line current. Otherwise referred to as "loop-disconnect" signalling or "pulse" signalling. It is the form of signals sent by an ordinary rotary telephone dial. It is now largely superseded by "tone" signalling (DTMF). At present all Spark TDM exchanges are capable of responding to decadic signalling. However, as these become progressively replaced, decadic signalling will cease to be available. Derived circuit: is a circuit which is provided by means other than a physical pair of wires from the telephone exchange to the customer's premises. Typical examples of this are circuits over fibre optic and wireless systems. Direct dialling-in (DDI): is the facility to allow incoming calls from the PSTN to be switched directly to a specified station (e.g. PABX extension) without operator assistance. Distinctive Alerts (DA): are the four different ringing cadences (DA1 to DA4) which allow multiple devices connected to the same line to respond to specific cadences while ignoring others 7

Is mainly used to allow facsimile machines to share phone lines with a telephone. The Facsimile machine is set to respond to DA4 and ignore the standard telephone ring cadence DA1 Double talk: relates to echo control and describes the condition whereby signals are present in both directions of a 4-wire circuit at the same time. This occurs when both parties in a telephone conversation are speaking at once and the situation requires special treatment by any echo control device present in the circuit. DTMF (Dual Tone Multi-Frequency): is a signalling system used over PSTN customer lines whereby two tones are sent simultaneously to line for each digit. It is used both for call initiation and for the accessing or controlling of other services, often between customers, following connection of a call. The DTMF standard is described in ITU-T Recommendation Q 23. Echo: is an unwanted signal reflection delayed to such a degree that it is perceived as distinct from the signal directly transmitted. In telecommunication networks, there is a distinction made between "talker echo" and "listener echo" as follows:- (a) "Talker echo" is the reflected signal experienced at the terminal sending the original signal. This is particularly disturbing to the person speaking in a telephone conversation who hears their own voice returning, but delayed enough to disrupt their flow of speech. (b) "Listener echo" is the reflected signal experienced at the terminal receiving the original signal. This can be a problem for data transmission since the receive terminal is likely to receive the same signal twice, the second being sufficiently delayed, but of high enough power level to be interpreted as another valid signal. Echo control device: is a device, operated by voice signals, which is used in telecommunication networks to reduce the effect of echo by either suppressing or cancelling the echo signal. The "Echo suppressor" was the earlier form of echo control used internationally. This device reduced the effect of echo by introducing additional loss in the echo path. It is rarely used now. Later technology developments introduced the "echo canceller" which is a more effective method of controlling echo. This estimates the echo signal from an examination of the original signal and subtracts that estimated signal from the actual echo signal without affecting the transmission path. External Test Point (ETP): is the terminal box, fitted at the customer s end of a cable lead-in, in which the lead-in cable is connected to the building cabling. The ETP may also be used to house a simple electrical termination which allows remote testing of the line from the exchange through to the customer s premises when no terminal equipment has been connected. Full current: is the current drawn by any item of terminal equipment when connected directly to a 50 V, 400 Ω source in the off-hook condition. "Full current" is used for test purposes and defines the maximum current that can be drawn under zero line conditions. Most lines are current limited to well below the current that a 50 V, 400 Ω source would deliver. Current limits as low as 25 ma are often used. 8

Individual line: is a line serving a single customer. An individual line may have one or more devices connected within that customer premises. Inter-digital pause: is the interval between successive DTMF tone bursts in a series of digits. Key telephone system (KTS): is a small telecommunications system designed for use in customer's premises which provides switching facilities between individual extension devices and the network connection. Simple KTS's often have a nominal 0 db (i.e. 1 db) transmission loss between ports. See also "PABX" for further details. For the purposes of this and other TNA documents, and PTC Specifications, the term "PABX" embraces all Private Automatic Branch Exchange (PABX), Key Telephone System (KTS), Small Business Exchange (SBX) and other equipment intended for installation in a customer's premises to switch calls between separate telephone lines and extensions. Line impedance: is the terminating impedance presented to a line by any equipment to which it is connected. Loop current: is the standing d.c. current drawn by any equipment in the off-hook condition. The loop current is dependent upon the resistance of the equipment, the line length and any current limiting by the exchange line feed equipment. Loudness Rating (LR): is a measure, expressed in decibels, for characterising the loudness performance of complete telephone connections, or parts thereof, such as the sending system, line, or receiving system. Reference ITU-T Rec. P. 64:1993, P. 65:1993, and CCITT Blue Book, Rec. P. 76. Loudness rating is an internationally accepted method of objectively measuring the performance of telephones from the mouthpiece to a given point on the line, and vice versa to the earpiece. The approach enables computer-controlled measuring equipment to be used for making quick, accurate and, above all, repeatable tests. A loudness value is the result of a calculation based on fourteen separate measurements made at predetermined frequencies within the normal telephony frequency range, each measurement being "weighted" according to its effect as perceived by the human ear when listening to normal spoken words. The loudness measurement value is actually the loss involved in the circuit under test, relative to an internationally accepted reference standard. Thus the higher the loudness value the quieter the perceived signal volume. A negative value occurs when the loss is actually less than that of the reference standard. Overall loudness rating (OLR) is the sum of the send loudness rating (SLR) of the telephone at one end of a telephone connection, the receive loudness rating (RLR) of the telephone at the other end, and the loudness ratings of each section of line in between. In other words it is a measure of the overall electro-acoustic performance between mouthpiece at one end and earpiece at the other. Master jackpoint: is a telecommunications outlet which provides the 'on-hook' line termination for a Telecom PSTN line and derives the 'shunt' wire for 3-wire connection. The 3-wire system for customer premises wiring was introduced in the 1980s as a means of preventing decadic phones causing bell tinkle in other parallel connected phones during dialling. It is no longer used although some older installations may still have it. It is detrimental to xdsl services and should be converted to 2-wire. 9

Off-hook: is the condition where the equipment is connected to line and is used to initiate or take part in a call. On-hook: is the condition where the equipment is connected to line in the idle state awaiting receipt of an incoming call or available to initiate a call. The above terms are derived from the term "hookswitch", which is used to describe any device which changes the status of the equipment from "on-hook" to "off-hook" or vice versa. On-hook data transmission: is the transmission of information in the form of data signals over a PSTN line while the terminating CPE is in the on-hook condition. This data is typically used for transmitting Caller ID and message waiting information to a customer. PABX (PBX): Private Automatic Branch Exchange (Private Branch Exchange) is a form of telecommunications system designed for use in a customer's premises which provides full switching facilities between individual extension devices and the network. See also "Key Telephone Systems". For the purposes of this and other TNA documents, and PTC Specifications, the term "PABX" embraces all Private Automatic Branch Exchange (PABX), Key Telephone System (KTS), Small Business Exchange (SBX) and other equipment intended for installation in a customer's premises to switch calls between separate telephone lines. For Telepermit purposes, it is necessary to divide PABX's/KTS's into two defined categories as follows:- (a) Type 1: 4-wire switching devices (digital or analogue) which, by their very nature are designed to have an inherent 2-3 db transmission loss between extension and 2-wire analogue trunk ports (ref. Specification PTC 109). (b) Type 2: 2-wire analogue switching devices without networking facilities and which have a nominal 0 db ( 1 db) transmission loss between extension and trunk ports. Most KTS's can be categorised as Type 2 and most large PABX's as Type 1, but this is not always the case. For the purposes of defining interface requirements a PABX system may be considered to provide a similar range of conditions to that of a public exchange line. Psophometric: is the term used to describe a method of measuring noise within the speech band while weighting the value of each frequency component present in accordance with its relative effect on the human ear. Such measurements are normally made with a psophometer, which is a voltmeter fitted with a standardised frequency weighting network and calibrated to indicate noise power or voltage in psophometric units (dbmp or mv psophometric). The weighting coefficients defined in CCITT Blue Book, Rec. O. 41 for telephone circuits and weighted to a reference tone of 800 Hz are used by Telecom for telephony purposes. PSTN: is the Public Switched Telephone Network. New Zealand PSTN services may be provided by a number of different Network Operators, each of which can set different network interface requirements should they choose to do so. Recall: is the procedure used to re-connect the register function of a switching system to enable additional features of that system to be used while a call is in progress. 10

Ringer (or ringing detector): is any device which responds to the alternating voltage applied to indicate an incoming call. Secondary jackpoint: is a telecommunications outlet which provides an additional connection point to a Telecom PSTN line which is also equipped with a Master jackpoint. This jackpoint provides only a CPE connection facility and is wired in parallel to a Master jackpoint. Service Delivery Point (SDP): is the defined electrical interface point provided at an agreed physical location to which Telecom will deliver service to a customer. In commercial premises, the SDP may or may not be the same as the network demarcation point. Cable owned by a third party, such as the building owner, may be used to serve the SDP. Signalling: is the exchange of information (other than by speech) which is specifically concerned with the establishment, release and other control of calls over the PSTN. The term "signal" can also be used in connection with other types of transfer of information, but only if suitably qualified, e.g. "data signal", "voice signal". Telecommunications outlet (TO): is any jackpoint forming part of the fixed wiring in a customer's premises at which CPE may be connected to a telecommunications network. This was typically a BT jackpoint, but is being superseded by the 4 pair RJ45 socket Telepermit: is the Registered Trade Mark used to indicate Spark s agreement to the connection of equipment to its network. Two-wire jackpoint: is a version of the BT jackpoint which superseded the Master/Secondary jackpoints. This in turn was later superseded by the 2-C (two contact) jackpoint which only had contacts 2 and 5 equipped to increase reliability by increasing the distance between contacts. 11

4 TRANSMISSION CHARACTERISTICS 4.0 General The PSTN network has been designed for switching voice calls, although has been also used for switching voice band data calls using modems. To get the maximum perceived dynamic range a simple compression algorithm is used. In New Zealand the codecs used are ITU-T G.711 A-law codecs. Compression is matched to the human perception of sound which is approximately logarithmic, so as the level increases, the quantization steps increase in size The networks uses two basic transmission techniques, Time Division Multiplex (TDM) and Packet based, increasingly using Internet Protocol (IP). Often an end to end Call will use a combination. 4.0.1 Time Division Multiplex (TDM) TDM was used in the first digital transmission systems, and was the dominant technology for voice networks up until the beginning of the 21 st century. In a conventional TDM network, the analogue voice signal is band limited to 3100 Hz (300 3400 Hz), then sampled at 8000 times a second, and converted by an A-law codec into an eight bit word. This gives a digital transmission rate of 8000 Samples/second X 8 bits per sample = 64,000 bits/sec. (64 kbps). At the other end of the circuit, the 64 kbps bitstream is converted back to an analogue signal 8 bits at a time. Every second, this conversion back to analogue will happen 8000 times. Within the voice bandwidth of 3100 Hz, the analogue signal reproduced at the far end will theoretically be an exact copy of the analogue signal at the sending end. The Multiplex part of TDM is simply a method of interleaving multiple 64 kbps voice streams, for example 32 64 kpbs circuits can interleaved to form one 2.048 Mbps circuit known as an E1 circuit. Likewise, multiple E1 circuits can be interleaved to form higher bit rate circuits. Strengths of TDM An end to end circuit is inherently synchronised, so the bits are clocked out of the circuit by a clock which is synchronised to the clock clocking the bits into the sending end of the circuit. This is essential for high speed data modems to operate. Once the end to end call is established, the transmission resource is guaranteed for the duration of the call. There is little or no processing resource required once the call is established. Relatively low delay makes it good for real time interactive communications such as voice or video conferencing. Security from hacking. As the end to end data is carried separately from the call control information, and the call control information is often carried on a physically separate network which is not accessible publicly, TDM networks are relatively secure. Weaknesses of TDM Inefficient for bursty data such as internet browsing where the full capacity of the circuit is only used occasionally. Inflexible. Because it has evolved specifically for the 64 kbps voice application it is difficult to adapt it for anything else such as wide bandwidth voice. 12

4.0.2 Packet transmission especially Internet Protocol (IP) Unlike TDM circuits which connect two end points for the full duration of a call, end to end IP communication is connectionless. For voice over IP telephony a packet is typically assembled from 160 x 8 bit voice samples, and the packet then sent to the distance end. Each packet contains all the information required by the network to route it to the far end. From the network perspective each packet is a self contained event, and on a packet by packet basis, the network will attempt to find enough resource. Whether it succeeds or not depends on how busy the network is at that instant. Strengths of IP Efficient use of resource Good for intermittent (bursty) data as resource is only used when necessary. Good for unidirectional non-time-critical data Weaknesses of IP Not end to end synchronised. This is not generally a problem, but voice-band data modems may be unreliable as the all modems above 2400bps use phase locked loops to synchronise their internal clocks. Jitter between end points makes this process unreliable. IP networks generally have larger end to end delays making them unsuitable for real time interactive communications such as voice and video conferencing. This is in part due to the fact that every packet has to be processed separately, and also due to an IPv4 packet header being 70 bytes long it becomes very inefficient to send short packets. Because network control information is carried in the same packets as user information, an IP is inherently less secure. While there are means of mitigating the security issues, these general increase the delay. 4.1 Loss Plan (1) The loss plan follows ITU-T Recommendations which specifies a Send Loudness Rating (SLR) from a telephone handset to the 0 dbr point of 8 db and a Receive Loudness Rating (RLR) from the 0 dbr point to the handset of 2 db. These are made up of several components shown in Figures 4.1 and 4.2. (2) If the telephone is connected via a local access cable, the loss plan assumes that the traffic weighted average loss in the local cable is 2.5 db, and the interface loss PADs are set at 0.5 db (T- Pad) and 6.0 db (R-Pad) as shown in Figure 4.1. The parameters given are at the MDF for a Telephone exchange or at the port on a derived circuit, so the frequency response and overall loss will be modified by the characteristics of the cable between the customer equipment and the network equipment port. (3) If the Interface is in the customer s premises the loss between the interface and the telephone may be assumed to be close to zero, so for consistency, the 2.5 db local cable loss is added to the T and R Pads in the customer located ATA as shown in Figure 4.2. 13

(4) Networks are interconnected with each other at the 0 dbr point. These interconnections are digital, so there is no loss between networks (both nationally and internationally). If private voice networks are connected to public networks digitally (ISDN or SIP trunking) they will usually extend the 0 dbr point into their own network. (5) The maximum level at the input to the codec is +3.14 dbm. This is equivalent to 3.14 Vpp (600 Ohm). On a zero-length line for a reticulated access connection, the maximum input level will be increased by 0.5 db so the maximum input level will be 3.64 dbm or 3.33 Vpp. When access is via a customer located ATA, then the maximum level at the port will be increased by an additional 2.5 db. The maximum input level at the ATA port will be 6.14 dbm (4.44 Vpp). SLR(to 0dBr point) = 5.0 + 2.5 + 0.5 = 8 db SLR = 5.0 db Traffic weighted mean local access cable loss = 2.5 db 4/2 wire hybrid Attenuation from the customer to the network T-Pad 0.5 db A D Balance Impedance BT3 Analogue Telephone complying with PTC200 RLR = -6.5 db SDP RLR(from 0dBr point) = -6.5 + 2.5 + 6 = 2 db R-Pad 6.0 db Attenuation from the network to the customer A D 0 dbr point Figure 4.1 Transmission losses for connection to line card via reticulated cable SLR(to 0dBr point) = 5.0 + 3.0 = 8 db SLR = 5.0 db 4/2 wire hybrid Attenuation from the customer to the network T-Pad 3.0 db A D Balance Impedance BT3 SDP RLR = -6.5 db Attenuation from the network to the customer RLR(from 0dBr point) = -6.5 + 8.5 = 2 db R-Pad 8.5 db A D 0 dbr point Figure 4.2 Transmission Losses for connection to customer located ATA 14

4.2 Frequency Response The frequency response of the network is nominally 300-3400 Hz. The objective frequency response is shown in Figure 3.3 1.70 Loss relative to loss at 1000 Hz (db) 1.5 1.0 0.75 0.70 0.5 0.35 0-0.3 Out of spec region 300 400 600 2400 3400 1000 2000 3000 4000 Frequency (Hz) -0.5 Out of spec region Figure 4.3 Frequency response of the analogue interface Notes: 1. This is an objective frequency response for a signal of -10 dbm0 between the analogue 2- wire interface and the 0 dbr point in both directions. 2. The frequency response is relative to the loss at 1000 Hz. 3. The objective frequency response assumes that the input and balance impedances are both BT3. 4. In practice, the frequency response at the SDP will vary according to the length of cable between the line interface and the SDP. Generally, as the line increases in length, the high frequencies will be attenuated more than the lower frequencies. 5. The frequency response is specified by the ITU-T in Recommendation Q.552. 15

4.2 Variation of Gain with Input level The loss/gain variation with input level is shown in Figure 4.4 1.6 1.5 1.0 0.6 0.5 Out of Spec region Gain Variation (db) 0.3 0-0.3-55 -50-45 -40-35 -30-25 -20-15 -10-5 0 +5 Input level (dbm0) +3-0.5-0.6 Out of Spec region -1.0-1.5-1.6 Figure 4.4 Variation of Gain with input level 4.3 Impedance 4.3.1 Input Impedance The impedance looking into the port is nominally BT3. This is a passive network made up of a 370 Ohm resister in series with a parallel combination of a 620 Ohm resistor and a 310 nanofarad Capacitor (see fig 4.5). The significance of the input impedance is that it must match the balance impedance of the customer equipment. Mismatch will increase the sidetone in the customer equipment to a point where the customer equipment can oscillate if the mismatch is bad enough. For a telephone, the sidetone is measured directly rather than the telephone balance impedance. 4.3.2 Network balance impedance The network balance impedance is also BT3 (see 4.3.1 and fig. 4.5). For optimum performance, the input impedance of the customer equipment must match the Network Balance impedance. Mismatch will cause signals to circulate within the network which is heard as echo at the other end of the call. If the network connection is short enough the echo will appear to be increased sidetone. See fig. 4.6. The degree of match between Customer Equipment Input Impedance and BT3 is measured as a return loss at frequencies across the voice band (300 to 3400 Hz). 16

370 Ohm 310 nf 620 Ohm Figure 4.5 BT3 Network 4.4 Network Echo Control 4.4.1 Causes of echo (1) The total control of echo by means impedance matching alone (ref. clause 4.3) is not practicable due to component tolerances and other variables. There is always a certain amount of signal reflection occurring whenever there is a 2-wire/4-wire transition. This particularly becomes a problem when signal delay is introduced in the transmission path, e.g. on long distance or international calls. This delay causes the signal reflection to become noticeable as 'talker echo' which can make telephone conversation extremely difficult. Further 'near end' reflections produce 'listener echo' which is often an additional problem for the transmission of data. (2) The change from analogue to digital transmission around the world during the 1980s and 1990s, further compounded the problem by introducing additional processing delays into transmission paths. The later use of packet based systems further increases delay requiring extensive use of echo control measures. 4.4.2 Control of echo (1) There are two types of echo control devices commonly used on circuits where echo is likely to result on PSTN calls. These are as follows: (a) Echo suppressors: These control echo by inserting a loss of at least 30 db in the transmission path in the direction opposite to that of the original signal. This loss is removed after a period of 100 ms of quiet time. When signals are present simultaneously in both directions for a period of 50 ms, the echo suppressors enter the "double talk" state in which 6 to 15 db of loss is added to both directions of transmission. Echo suppressors were used in the 1980s, but are no longer deployed as they are inferior in performance to echo cancellers. It is unlikely that echo suppressors will be encountered in the current network. (b) Echo cancellers: These control echo by adding loss to the echo signal only, without affecting the transmission path. After a training (converging) period of approximately 500 ms, the echo canceller will assure an echo loss of at least 40 db. During the "double talk" state the degree of echo cancellation may diminish slightly, but the transmission path is again not affected. At the start of a call echo cancellers characterise the call path and measure the levels and delays of the echo signals which are then subtracted from the signal thus eliminating the echo. To further reduce the low-level echoes, a non-linear processor is used to attenuate signals below a noise threshold. 17

P1 P2 6.0 db 0.5 dbp3 P4 ZB(A) ZIN(A) ZIN = BT3 Local 2-wire Access Cable 2.5 db loss ZB = BT3 ZB = BT3 ZIN = BT3 ZIN(B) Local 2-wire Access Cable 2.5 db loss ZB(B) Telephone (A) Telephone (B) 0.5 db 6.0 db PSTN P1 is the component of the send power which is returned to the receiver as a result of mismatch between the input impedance of the PSTN (BT3) and the Balance Impedance of Telephone (A). It is perceived as sidetone. P2 is the power reflected back towards Telephone (A) from the network interface due to the mismatch between the input impedance of Telephone (A) and the network Balance impedance (BT3). Due to the short propagation delay between Telephone (A) and the network interface this is also perceived as sidetone. If there were significant delay between Telephone (A) and the network. This component would start to appear as echo. P3 is the component of send power (from Telephone (A)) which is sent back to Telephone (A). It is due to the mismatch between the input impedance of Telephone (B) and the balance impedance of the network at the Telephone (B) interface (BT3). It is perceived at Telephone (A) as increased sidetone if the network delay is small, or echo if the network delay is large (eg and international call). P4 is the component of send power from Telephone (A) which is reflected back towards Telephone (A) as a result of a mismatch between the balance impedance of Telephone (B) and the input impedance of the network interface (BT3). It is perceived at Telephone (A) as increased sidetone if the network delay is small, or echo if there is significant network delay. Notes: Every time there is a mismatch there is not only an increase in sidetone or echo, but also the power received by Telephone (B) is reduced by the same amount. The same degradations occur at Telephone (B) for signals transmitted to Telephone (A). An Echo signal towards Telephone (A) is the same as a send signal from Telephone (B) and itself may be audibly echoed back to Telephone (B). This is known as listener Echo as against Talker Echo. Where delays are significant echo cancellers are used to eliminate echo, but cannot restore the levels which are reduced by impedance mismatch. Echo cancellers also produce distortion which makes them unsuitable for circuits carrying facsimile or other voice band data traffic. FIGURE 4.6 End to end network schematic showing the effects of impedance mismatch on network performance. 18

4.4.3 Disabling of echo control devices Echo control devices are designed specifically to improve telephone conversations. However, on some connections, they can cause greater problems than echo itself particularly the non-linear processor which can enhance voice by removing noise, but may destroy signals carrying valid data. These are typically connections between modems and facsimile machines. It is therefore necessary that echo control devices are capable of being disabled when required. For this reason, they are equipped with tone operated disablers designed to respond to a frequency of 2100 Hz. To disable the echo canceller the terminal at either end of the connection shall transmit a signal with the following characteristics: Frequency: 2100 +/- 15 Hz Phase change: 180 degrees +/- 25 degrees every 450 +/- 25 ms (the phase must have changed by 180 +/- 10 degrees within a period of 1 ms. During the phase change the level of the signal shall not drop more than 3 db below the steady state value for more than 400 ms Level: -31 dbm0 to 0 dbm0 (recommended -15 dbm0) Duration 3.3 +/- 0.7 seconds Out of band power: Signals other than the 2100 Hz shall be at least 15 db below the 2100 Hz signal. Distant end terminal: The terminal not transmitting the echo canceller disabling signal shall not transmit any signals above -46 dbm in the 200 to 4000 Hz band. Note that transmitting the 2100 Hz tone without the phase reversals will not disable the echo cancelling process but should disable the non-linear processor only, which will remove some distortion which a modem is likely to be sensitive to. Most of the higher speed modems have their own echo cancellers so the network canceller is not required. 4.4.4 Holding the disabled condition The tone detector will hold the echo canceller and non-linear processor in the disabled state for any single frequency sinusoid in the band from 390 700 Hz having a level of -27 dbm0 or greater and from 700 3000 Hz having a level of -31 dbm0 or greater. The disabler will release where the signal in the band 200 3400 Hz drops below 36 dbm0 for more than 250 ms. 4.5 Balance about earth The 2-wire network interface is balanced about earth, and to keep induced noise to a minimum it is important that customer equipment and cabling maintains this balance as much as possible. The Balance about earth requirements for Customer equipment are documented in the applicable PTC Specifications. For CPE, Balance about earth across the voice band of 40 db is a minimum requirement with 60 db being recommended. 19

5 SIGNALLING 5.1 Signalling types (1) The standard method of signalling between customer premises equipment and Spark analogue network interfaces is dual tone multi-frequency (DTMF) signalling. This method is also widely used for signalling from customer to customer after a call has been established. Reference ITU-T, Recommendation Q. 23. (2) (a) Historically, the standard method of signalling was decadic, consisting of trains of break pulses to indicate the digits signalled. The version of this system developed for use in New Zealand was unique in that the coding was the reverse of that used generally around the world. That is, when the number N is dialled, the number of pulses sent to line is 10 N. (b) The use of decadic signalling will not be available on any of the new network interfaces, and will only remain available on legacy TDM networks until it is phased out. (3) Direct Dial-in (or DDI ) operation is a particular type of signalling used between the exchange and the customer s equipment whereby the final 1-4 digits of the called number are sent from the exchange. DDI operation can use decadic or DTMF signalling, as required by the customer s equipment. Further details of Analogue 2-wire DDI can found in PTC 107, although it should be noted that this type of interface will be phased out as legacy TDM exchange equipment is retired from service. 5.2 DTMF signalling 5.2.1 DTMF tones (1) The allocation of DTMF signalling frequencies necessary to signal information to the network is as follows:- LOW GROUP (Hz) High Group (Hz) 1209 1336 1477 1633 697 1 2 3 A 770 4 5 6 B 852 7 8 9 C 941 * 0 # D The 'A', 'B', 'C' and 'D' signals are not currently used by the network but may be used in ened to end communications. 5.2.2 DTMF Characteristics The DTMF receiver will respond to DTMF signals in the following ranges:- (a) (b) Any receive level between -5 dbm and -20 dbm. High frequency pre-emphasis of between 0 and 3 db. (c) DTMF frequencies within ±1.8 % of the nominal values (ref. PTC200 clause 5.2.1(1)). 20

(d) The receiver will recognise any valid DTMF signal that is present for a minimum of 55 ms, as long as it is preceded by a continuous pause of 55 ms. (e) The receiver will ignore breaks of up to 15 ms provided the signal either side of the break represents the same digit, and the break does not occur within 20 ms of the start or the finish of the tone burst. The following DTMF signals will be rejected (a) (b) A signal of less than 20 ms duration A signal of less than -40 dbm (c) A signal in which either of the individual frequencies deviates by more than +/- 3.5% of the nominal frequencies listed in ITU-T recommendation Q.24. (d) Signals where any frequencies other than the correct DTMF pair are also present shall be rejected as valid DTMF if the total power of such frequencies is greater than the level of the lowest power valid frequency minus 20 dbm. 5.2.3 DTMF signalling between customers The received level of DTMF tones at the customer's premises when transmitted from a distant point in the PSTN, will on average be in the order of -20 dbm, but may be as low as -40 dbm depending on actual transmit level and length of line. If transmitted from other networks, it may in some extreme cases be even lower. 5.3 Recall (also known as timed break recall (TBR) or Hookswitch flash (HSF)) This is used to activate special network features and is invoked by breaking the d.c. loop for a short period. The length of the break must be short enough not to terminate the call and long enough to be recognised. The network is designed to respond to range 500 to 800 ms, but will usually respond to times down to 350 ms. 21

6 D.C. LINE CONDITIONS 6.1 General The d.c. conditions at customers' demarcation points on PSTN lines vary depending on the following:- (a) The type of line feed equipment used, and (b) The length of the line and type of cable used. 6.2 Exchange line feed equipment (1) The nominal supply voltage used in local exchanges to feed PSTN lines is in the range 50 V to 90 V. However, for most lines a nominal 50 V power supply is used, having a tolerance range of 44 V to 56 V. The positive pole of the power supply is connected to earth (see Fig. 6.1). Where the analogue interface is derived over fibre or wireless, there may be no connection to earth, i.e. the interface is floating. The exchange supply may be boosted to 90 V between the line wires on some long rural lines, but the voltage on each conductor will not normally exceed ± 52 V with respect to earth. (2) The line current supplied will normally be current limited. The limits will depend on the type line circuit used but can range from 80 ma down to 20 ma. The low current limit feeds may be referred to as constant current feeds as in most cases they will be operating in the current limited mode. (3) Ripple components up to 2 mv psophometric may be present. (4) Under normal network switching conditions step changes in the open-circuit voltage may occur within the limits stated. A typical case of this is the "reversal on answer" supervisory signal. Other network conditions also result in short breaks in line current during the setting up of a call. 22

_ + Current limit 20-80 ma 50 V 200 Ohm Hook Switch d.c. load (hold circuit) 200 Ohm Access Cable Customer Equipment Network Interface Service Delivery Point Note: The 200 Ohm resistors represent the d.c. impedance. At voice frequencies the a.c. impedance will be generally > 50,000 Ohm. Historically this was achieved by the use of large inductors, but is now implemented using an electronic gyrator circuit. FIG. 6.1 EXCHANGE LINE FEED 6.3 Derived circuits (1) Use of derived circuit equipment for all or part of the line between the core network and the customer, is widespread and on the increase. In such cases, the line feed current is produced an equipment terminal nearer to the customer. (2) The analogue interface may be derived in a street side cabinet from a fibre feed, and reticulated to the end customer over conventional copper cable, or the fibre may go into the customer premises with the derived analogue circuit being connected directly into the premises wiring. typically derived circuits provide lower current d.c. feeds than the old telephone exchange feeds. 6.4 Line polarity The polarity of the pair at the Service Delivery Point is not fixed and may change at any time following network maintenance. 6.5 Answer supervision Answer supervision is provided on originated calls by means of a reversal of the line polarity when the called line is answered. The line polarity is again reversed should the called party be the first to release the call. Line polarity reversal for answer supervision does not apply on the called party s line should the calling party release first. In this case, the only indication given to the called party is disconnect tone. 23

6.6 Clear Forward Clear forward is indicated by a break of 800 to 1100 ms in the positive lead when the distant party terminates the call. Answer supervision and Clear Forward are NOT part of the standard PSTN service and are only available on business lines on request. They may not be available on all lines. 6.7 Voltage transients (1) Changes to line conditions of up to 50 ms duration (for example, polarity, voltage, and feeding resistances) may occur during processing of a call by the network or by PABX's. (2) Also, high voltages may occur in the event of power system faults or lightning strikes. 6.8 Line Tests Maintenance staff may employ 500 V insulation resistance measuring equipment when testing lines connected to the PSTN. Such voltages will not be encountered on ATA derived circuits. 6.9 Requirements for terminal equipment For various states, the following conditions apply:- Line Condition Seize line Hold line Answer line Release line Hookswitch flash Loop Current-time >15 ma for > 10 ms >15 ma indefinite or time out after 10 seconds >15 ma for t > 40 ms < 5 ma for t > 1000 ms < 5 ma for 500 ms < t < 800 ms 24

7 RINGING CHARACTERISTICS 7.1 Ringing frequency The standard ringing frequency applied to PSTN lines is 25 Hz. 7.2 Ringing voltage (1) For individual Spark PSTN lines, the conditions are as follows (see also Fig. 7.1):- (a) The open circuit ringing voltage applied to line at the exchange is nominally 90 Vrms, dropping to 75 Vrms under the maximum rated load of the ringing generator. (b) One side of the ringing generator is normally connected to earth at the exchange, and the other applied to one conductor of the line. The return path is via the other conductor to the exchange -50 V supply, the positive side of which is also connected to earth (see Fig. 7.1). (2) For Spark PSTN service using derived circuit equipment, the conditions are:- (a) On some derived circuits the ringing voltage may be somewhat less than in sub-clause (1) above, and the current available may be limited. (b) The ringing voltage at the customer's premises should still be at least 10 Vrms. (c) The ringing on derived circuits may be balanced about earth or not referenced to earth at all. For derived circuit terminal equipment, the maximum distance between the ringing supply and the customer is normally less than in (1) above, so the higher source voltage is not necessary. 7.3 Ringing current The ringing current available at the customer's premises is limited by the source impedance, the line impedance and the series capacitor at the line interface. To ensure reasonable ringing performance, the total ringing load connected to a line should not exceed a RN of 5. See Specification PTC 200 for details. 25

1100 Ohm 1 microfarad Capacitor 80 Vrms Ringer 50 V Access Cable Customer Equipment Network Interface Service Delivery Point Notes: Where service is provided via an ATA, the network interface and Service Delivery Point are effectively the same. Where the network is delivered to the Customer Premises via Chorus Cable, there will be some loss in the Access Cable, although at 25 Hz this will be largely resistive, at approximately 168 Ohms per km. A ringing load of RN = 1 (typical for a telephone) is the equivalent of a 1 microfarad capacitor in series with an 8 kohm resistor. For a modem or facsimile machine the load is usually much lower than for a telephone as it needs only to sense the presence of ringing and not provide power for an audible ringer. FIG. 7.1 RINGING CONNECTIONS 7.4 Ringing cadences (1) There are four distinctive cadences available for use on individual customers lines, and these are designated as Distinctive Alerts 1-4 (DA 1-4). The purpose of using different cadences is to enable an individual line to be used for up to four separate functions (e.g. telephone, fax, modem and answerphone, can all be supported by the one line). Use of suitable decoding arrangements associated with the CPE allows the nature of the incoming call to be determined before it is answered and avoids the ringing of all devices unnecessarily. All ringing cadences may not be available on all connections. DA1 is the standard cadence and will always be available. 26